Referring to FIG. 1, multi-capillary viscometers can introduce some type of delay unit in order to make a differential measurement while a sample is being measured. An illustrative prior art four-capillary viscometer 10, for example, includes four pieces of tubing or capillaries connected in a series-parallel configuration to form the hydraulic equivalent of a Wheatstone-Bridge in electronics. These tubing or capillaries in the arms of the bridge are often referred to as R1, R2, R3, and R4 because they are in effect hydraulic resistors. The delay unit 12 is placed in series with one of the capillaries and usually it consists of a column packed with or containing a material or solvent that will delay the sample from reaching a reference capillary while a measurement is taking place. This delay unit should generally provide for sufficient time or volume to accommodate the entire elution volume of the analytical GPC (Gel Permeation Chromatography) column set that is used for the separation analysis. In liquid chromatography there are a vast number of column set configurations, many requiring different delay volumes. Initially, the capillary configuration is arranged such that the “bridge” is “balanced” meaning that the DP+ & DP-− readings are approximately equal.
In the illustrative viscometer, R1, R2, R3, and R4 are capillary tubes of a small diameter giving them a measurable resistance to the solvent flow, and if R1=R2=R3=R4, the differential pressure (DP) output should theoretically be zero. This is the output signal from the bridge and should be within a small percentage of the total pressure across the bridge measured between the two differential measurement points IP+ and IP− when solvent is flowing. This is called the bridge balance and is given by the equation, Balance=4DP/IP−2DP, where DP is the differential signal from the DP+ and DP− readings measured in Pascals and IP is as mentioned above measured in Pascals.
Capillary bridge viscometers are described in more detail, for example, in U.S. Pat. No. 4,463,598 to Haney, which is herein incorporated by reference.
When a delay volume is placed in series with one or more capillaries, the bridge can be balanced or rebalanced to make up for additional resistance introduced by the presence of the delay volume(s). This can be accomplished by adjusting the length(s) of one or more of the capillary tubing runs to get the bridge balance back to the manufacturing standard balance. Upon installation or during use, it may become necessary to adjust this delay volume according to the analytical column set required for analysis. One or more additional delay volumes of different sizes may therefore be shipped with the instrument or purchased to meet the specific need of the customer. With these changes comes either increased or decreased resistance within the combined capillary and delay column flow path, and the viscometer can be rebalanced by adding or subtracting to the length of the appropriate capillary tubing in order to achieve the most efficient performance by returning to a balanced condition.
The traditional method for balancing a viscometer bridge is to change the length of one or more of the capillary flow paths. This is accomplished by calculating the amount to subtract (or add) from a length of one or more of the capillaries. The bridge is then disassembled to make the change and reassembled by a skilled technician. This can be extremely inconvenient and may also require the instrument to be returned to the manufacturer for qualified servicing. It is also common for the balance to change due to the introduction of different solvents. These changes are typically ignored because of the inconvenience and because the length difference involved can be physically too small to allow an accurate adjustment to be accurately accomplished, and the result can be a decrease in instrument performance.
Temperature-based balancing has also been proposed in U.S. Pat. No. 7,213,439 to Trainoff, which is herein incorporated by reference. But this approach can have the potential drawback of causing thermally induced changes to properties of the fluids in the viscometer. Understanding whether such changes are a concern for a particular experimental setup and whether they should possibly be compensated for can introduce the prospect of an undesirable level of theoretical complexity for the end user of the viscometer.